Synthesis of the Novel PARP-1 Inhibitor AG-690/11026014 and Its

Original Article

Synthesis of the novel PARP-1 inhibitor AG- 690/11026014 and its protective effects on angiotensin II-induced mouse cardiac remodeling

Guo-shuai FENG1, Cui-ge ZHU2, Zhuo-ming LI1, Pan-xia WANG1, Yi HUANG1, Min LIU3, Ping HE1, Lan-lan LOU2, Shao-rui CHEN1, *, Pei-qing LIU1, *

1Department of Pharmacology and Toxicology, School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou 510006, China; 2Department of Medicinal Chemistry, School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou 510006, China; 3Department of Pharmacology, Shandong University School of Medicine, Ji-nan 250012, China

Abstract We previously identified AG-690/11026014 (6014) as a novel poly(ADP-ribose) polymerase-1 (PARP-1) inhibitor that effectively prevented angiotensin II (Ang II)-induced cardiomyocyte hypertrophy. In the present study, we reported a new synthesis route for 6014, and investigated its protective effects on Ang II-induced cardiac remodeling and cardiac dysfunction and the underlying mechanisms in mice. We designed a new synthesis route to obtain a sufficient quantity of 6014 for thisin vivo study. C57BL/6J mice were infused with Ang II and treated with 6014 (10, 30, 90 mg·kg-1·d-1, ig) for 4 weeks. Then two-dimensional echocardiography was performed to assess the cardiac function and structure. Histological changes of the hearts were examined with HE staining and Masson’s trichrome staining. The protein expression was evaluated by Western blot, immunohistochemistry and immunofluorescence assays. The activities of sirtuin-1 (SIRT-1) and the content of NAD+ were detected with the corresponding test kits. Treatment with 6014 dose- dependently improved cardiac function, including LVEF, CO and SV and reversed the changes of cardiac structure in Ang II-infused mice: it significantly ameliorated Ang II-induced cardiac hypertrophy evidenced by attenuating the enlargement of cardiomyocytes, decreased HW/BW and LVW/BW, and decreased expression of hypertrophic markers ANF, BNP and β-MHC; it also prevented Ang II-induced cardiac fibrosis, as implied by the decrease in excess accumulation of extracellular matrix (ECM) components collagen I, collagen III and FN. Further studies revealed that treatment with 6014 did not affect the expression levels of PARP-1, but dose-dependently inhibited the activity of PARP-1 and subsequently restored the activity of SIRT-1 in heart tissues due to the decreased consumption of NAD+ and attenuated Poly-ADP-ribosylation (PARylation) of SIRT-1. In conclusion, the novel PARP-1 inhibitor 6014 effectively protects mice against AngII-induced cardiac remodeling and improves cardiac function. Thus, 6014 might be a potential therapeutic agent for heart diseases..

Keywords: cardiac remodeling; PARP-1 inhibitor; 6014; PARylation; NAD+; SIRT-1

Acta Pharmacologica Sinica (2017) 38: 638–650; doi: 10.1038/aps.2016.159; published online 27 Feb 2017

Introduction an important component of the RAAS, could induce cardiac [2–6] Cardiac remodeling, manifested clinically as hypertrophy remodeling following sustained hypertension . (enlargement), fibrosis, and dysfunction of the heart resulting PARylation, induced by poly(ADP-ribose) polymerases from cardiac load or injury, is generally accepted as a deter- (PARPs), is a type of reversible posttranslational modifica - minant of the pathogenesis of heart failure (HF)[1]. The renin- tion impacting numerous cellular processes, eg, DNA repair, [7] angiotensin-aldosterone system (RAAS) plays a pivotal role in transcription, metabolism, or immune functions . PARP-1, the physiological and pathological processes of cardiac remod- the predominant member of the PARPs, accounts for approxi- [8] eling. It is also widely accepted that angiotensin II (Ang II), mately 90% of the cellular PARylation events . Once activated, PARP-1 catalyzes protein PARylation, which results in the formation of polymerized ADP-ribose (PAR) dependently *To whom correspondence should be addressed. on NAD+ donor molecules, and subsequently the covalent E-mail [email protected] (Shao-rui CHEN); [9] [email protected] (Pei-qing LIU) attachment of PAR polymers to target proteins . Received 2016-09-26 Accepted 2016-11-24 PARP-1 is involved in pathological cardiac remodeling www.chinaphar.com Feng GS et al 639

and heart failure. Inhibition of PARP-1 prevents pathologi- eluted with a gradient of 5% to 95% solvent A for 25 min and cal cardiac remodeling induced by partial abdominal aortic detected at 254 nm; solvent A was CH3CN in double-deionized [10] 1 13 constriction . PARP-1-deficient mice are protected from Ang H2O. H NMR and C NMR spectra (AvanceIII, Burker) were II-induced cardiac hypertrophy[11]. Moreover, PARP-1 activa- recorded using TMS as an internal standard with a Burker - tion facilitates the pathogenesis of cardiac fibrosis[12] and the BioSpinUltrashield 400 NMR system. High-resolution mass development of heart failure[13, 14]. These findings indicate that spectra (HRMS) (LCMS-IT-TOF, Shimadzu) were recorded on PARP-1 could serve as a promising target for cardiac diseases, a Shimadzu LCMS-IT-TOF Instrument. and the pharmacologic inhibition of PARP-1 may be a potential therapeutic strategy[15–17]. General procedure for the synthesis of 6014 A previous study from our laboratory identified AG-690/ (1) 4-chlorophenol (77.8 mmol), 2-(chloromethyl)oxirane (306.9

11026014(6014) as a novel PARP-1 inhibitor from the Specs mmol) and K2CO3 (231.5 mmol) were refluxed in acetone database (www.specs.net), which contains approximately (100 mL) overnight, cooled to room temperature and filtered. 200,000 compounds based on virtual screening. We also The filtrate was concentrated and extracted with toluene and reported that 6014 could significantly attenuate Ang II- water, and then the 2-((4-chlorophenoxy)methyl)oxirane (M1) induced cardiomyocyte hypertrophy[18]. To further deter - in toluene was purified by column chromatography. mine the in vivo cardiac protective effect of 6014, the present (2) 3-methyl-1H-purine-2,6(3H,7H)-dione (30 mmol) and study used C57BL/6J mice infused with Ang II as the model, AcONa (60 mmol) were dissolved in AcOH, and Br 2 was and elucidated the effects and underlying mechanisms of added dropwise and reacted for 2 h at 65 °C. 8-Bromo-3- this novel PARP-1 inhibitor. This study aimed to determine methyl-1H-purine-2,6(3H,7H)-dione (M2) was generated by whether 6014, as a PARP-1 inhibitor, has therapeutic potential cooling, filtering, water washing and drying in a vacuum. for cardiac remodeling. (3) M1 (31.0 mmol) and M2 (25.8 mmol) in ethanol (120 mL) were combined with trimethylamine (2.6 mmol) and refluxed Materials and methods for 3 h. 8-Bromo-7-(3-(4-chlorophenoxy)-2-hydroxypropyl)- Reagents 3-methyl-1H-purine-2,6(3H,7H)-dione (M3) was produced by The compounds 3-aminobenzamide (3-AB), 4’,6-diamidino- cooling, filtering, ethanol washing and drying in a vacuum. 2-phenylindole (DAPI) and Ang II were purchased from (4) M3 (8.15 mmol) and 2-mercaptoacetic acid (16.3 mmol) in

Sigma-Aldrich (St Louis, MO, USA). Anti-PARP-1, anti- DMF (35 mL) were combined with NaHCO3 (32.6 mmol) and collagen I and anti-collagen III antibodies were provided reacted for 2 h at 85 °C. The reaction was stopped with water by Proteintech Group (Chicago, IL, USA). Anti-ANF and and then washed by methyl tertiary butyl ether. The products anti-FN antibodies were purchased from Santa Cruz Bio - in the water were combined with HCl (2.0 mmol) until the pH technology (Santa Cruz, CA, USA). Anti-BNP antibody was reached 3, and 2-((7-(3-(4-chlorophenoxy)-2-hydroxypropyl)-3- obtained from Merck Millipore (Bedford, Ohio, USA). Anti- methyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-8-yl)thio)acetic PAR antibody was obtained by Trevigen (Gaithersburg, MD, acid (M4) was yielded after filtering, water washing and dry- USA). Anti-SIRT-1, anti-lamin B1, anti-GAPDH antibodies ing in a vacuum. and all secondary antibodies for Western blot experiments (5) M4 (7.26 mmol) and ammonium chloride (21.7 mmol) were purchased from Cell Signaling Technology (Boston, in DMF (40 mL) was added to HATU (8.71 mmol) and DIEA MA, USA). An hematoxylin and eosin (HE) staining kit was (21.8 mmol) on ice and reacted overnight at room tempera- purchased from Santa Cruz Biotechnology (Santa Cruz, CA, ture. The products were precipitated by adding water and USA). A Masson’s staining kit and immunohistochemistry then filtering. 2-((7-(3-(4-chlorophenoxy)-2-hydroxypropyl)- reagents including Cy3-conjugated goat anti-mouse IgG (H+L) 3-methyl-2,6-dioxo-2,3,6,7-tetrahydro-1H-purin-8-yl)thio) were purchased from Wuhan Goodbio Technology Co, Ltd acetamide (6014) was purified by preparative chromatography (Wuhan, China). RIPA lysis buffer and IP lysis buffer were (Varian, Agilent). purchased from Beyotime Biotechnology (Jiangsu, China). The The synthesized intermediates M1-M4 were structurally protease and phosphatase inhibitor mini tablets, BCA protein identified by 1H NMR or MS, and 6014 was structurally identi- assay kit, Bradford protein assay kit, and protein G-agarose fied by 1H NMR, 13C NMR and HRMS. beads were obtained from Pierce Biotechnology (Rockford, AL, USA). A nuclear extract kit was purchased from ACTIVE 2-((4-chlorophenoxy)methyl)oxirane (M1) 1 MOTIF (Carlsbad, NM, USA). The quick start Bradford dye Crystal oil, yield 62.0%; H NMR (400 MHz, CDCl3) δ 7.28–7.22 reagent was obtained from Bio-Rad (Richmond, VA, USA). A (m, 2H), 6.90–6.83 (m, 2H), 4.23 (dd, J=11.0, 3.0 Hz, 1H), 3.93 SIRT-1 deacetylase fluorometric assay kit was purchased from (dd, J=11.0, 5.7 Hz, 1H), 3.39–3.32 (m, 1H), 2.92 (t, J=4.5 Hz, CycLex Ltd (Nagano, Japan). 1H), 2.76 (dd, J=4.8, 2.6 Hz, 1H). The inhibitor 6014 was synthesized in our lab. The purities of 6014 used for biological evaluation (99.61%) were deter - 8-bromo-3-methyl-1H-purine-2,6(3H,7H)-dione (M2) mined on a DIONEX Ultimate 3000 HPLC system (Chrome- White solid, yield 95%; 1H NMR (400 MHz, DMSO) δ 14.29 leon SR9 Build 2673): column, Acclaim 120 C18, 5 mm, 4.6 (s, 1H), 11.18 (s, 1H), 3.32 (s, 3H). MS (ESI, neg ion) m/z: 242.9 mm×250 mm; flow rate, 1 mL/min. The samples were [M-H]–

Acta Pharmacologica Sinica www.nature.com/aps 640 Feng GS et al

8-bromo-7-(3-(4-chlorophenoxy)-2-hydroxypropyl)-3-methyl-1H- diac output (CO), left ventricular fractional shortening (LVFS), purine-2,6(3H,7H)-dione (M3) left ventricular ejection fraction (LVEF), heart rate (HR), stroke White solid, yield 85.9%; MS (ESI, neg ion) m/z: 427.0 [M-H]– volume (SV), left ventricular internal dimension (LVID) in diastole and systole (LVIDd and LVIDs, respectively), left ven- 2-((7-(3-(4-chlorophenoxy)-2-hydroxypropyl)-3-methyl-2,6-dioxo- tricular posterior wall (LVPW) thickness in diastole and sys- 2,3,6,7-tetrahydro-1H-purin-8-yl)thio)acetic acid (M4) tole (LVPWd and LVPWs, respectively), and the interventricu- White solid, yield 90.5%; MS (ESI, neg ion) m/z: 439.0 [M-H]– lar septum (IVS) dimension in diastole and systole (IVSd and IVSs, respectively). All the parameters were obtained from 2-((7-(3-(4-chlorophenoxy)-2-hydroxypropyl)-3-methyl-2,6-dioxo- three cardiac cycles, and the data were analyzed by GraphPad 2,3,6,7-tetrahydro-1H-purin-8-yl)thio)acetamide (6014) Prism 5 (GraphPad Inc, La Jolla, CA, USA). 1 White solid, yield 30%; HPLC tR=11.24 min; H NMR (400 MHz, DMSO) δ 11.11 (s, 1H), 7.65 (s, 1H), 7.32 (d, J=8.7 Hz, Collection of tissues 2H), 7.23 (s, 1H), 6.93 (d, J=8.7 Hz, 2H), 5.52 (d, J=4.9 Hz, 1H), After the echocardiography measurements were completed, 4.37 (d, J=9.7 Hz, 1H), 4.27–4.15 (m, 2H), 4.01 (dd, J=9.9, 3.3 all the mice were weighed and anesthetized by sodium pen- Hz, 1H), 3.98–3.91 (m, 3H). 13C NMR (101 MHz, DMSO) δ tobarbital (40 mg/kg). The hearts of the mice were rapidly 168.66, 157.24, 153.93, 150.60, 150.53, 149.58, 129.14, 124.36, removed, imaged by photography, and weighed. Then, the 116.26, 108.11, 70.33, 67.65, 48.91, 36.68, 28.44. HRMS calcd. left ventricle was carefully trimmed and weighed. The ratio + for C17H18N5O5SCl [M+H] : 440.0790, found 440.0763. of heart weight to body weight (heart weight/body weight, HW/BW) and left ventricular weight to body weight (left ven- Animals tricular weight/body weight, LVW/BW) were analyzed. Healthy adult male C57BL/6J mice (10 weeks old, 20–25 g Half of the hearts from each group were stored in formalin weight) were raised in the animal center of Sun Yat-Sen Uni- and then embedded in 4% paraformaldehyde for later histo- versity and kept under pathogen-free conditions at approxi- pathological analysis. For the other half, the left ventricles mately 22 °C. All experimental procedures were conducted in were isolated and weighed. The left ventricles of the heart accordance with the Guidelines of Animal Experiments from were snap frozen in liquid nitrogen for Western blot analysis, the Ethical Committee for Animal Research of Sun Yat-sen SIRT-1 activity measurement and NAD+ content detection. University (No 44008500011624). Histopathological analysis Minipump implantation and administration For immunohistochemical staining, the tissues were fixed in As a commonly used PARP-1 inhibitor [19, 20], 3-aminobenza- 4% formalin and embedded in paraffin. Sections (5 mm) were mide (3-AB) was used as a positive control in this study. The cut from the embedded blocks and floated onto a warm (42 °C) C57BL/6J male mice were randomly assigned to 6 groups, water bath from where they were placed on Superfrost plus n=12/group: ① sham group; ② Ang II group; ③ Ang II+3-AB slides. The sections were deparaffinized and hydrated, and

(30); ④ Ang II+6014 (10); ⑤ Ang II+6014 (30); ⑥ Ang II+6014 the endogenous peroxidase activity was blocked with 3% H2O2 (90). 3-AB (30), 3-AB (30 mg ·kg-1·d-1); 6014 (10), 6014 (10 in water for 10 min. Antigen retrieval was performed with 10 mg·kg-1·d-1); 6014 (30), 6014 (30 mg·kg-1·d-1); 6014 (90), 6014 mmol/L citrate buffer (pH 6.0) for 10 min. The slides were (90 mg·kg-1·d-1). incubated with blocking reagent (Thermo Scientific) for 10 min Mice from all groups except the sham group were infused to block nonspecific binding. Then, they were incubated with with Ang II (1000 ng·kg-1·min-1) via subcutaneously implanted anti-PAR antibodies overnight at 4 °C, and later with HRP- minipumps (Alzet® pump model 2004, Cupertino, USA) for 4 conjugated secondary antibodies for 30 min at room tempera- weeks. Control mice were infused with normal saline. After ture. Finally, the slides were incubated with diaminobenzi- the minipumps were implanted, administration of 3-AB and dine and counterstained with HE. 6014 were carried out by intragastric administration twice a For HE staining, after deparaffinization and hydration, the day, and both 3-AB and 6014 were suspended in sodium car- sections were stained with HE. For Masson’s staining, after boxymethylcellulose. deparaffinization and hydration, the sections were stained with hematoxylin and Ponceau, as well as with phosphomo- Echocardiography lybdic acid and aniline blue successively in Masson’s staining. At the end of the experiment, all the mice were anesthetized For immunofluorescence, the frozen sections were fixed by with 1.5% isoflurane in oxygen (1 L/min). Two-dimensionally acetone. Bovine serum albumin (BSA) was used to block non- guided M-mode echocardiography (ESAOTE, Italy) was per- specific binding after antigen retrieval. The tissue sections formed using a Technos MPX ultrasound system equipped were incubated with anti-PAR antibody (1:100) overnight at with a 21-MHz imaging transducer, as described previously[21]. 4 °C and incubated with Cy3-conjugated goat anti-mouse IgG Ultrasound images of the heart were obtained by gently apply- (H+L) for 1 h. Subsequently, the nucleus was stained with ing the probe to the chest. The M-mode images were obtained DAPI. All the stained sections were sealed with neutral resin from the parasternal short-axis view at the papillary muscle. and photographed under an inverted fluorescence microscope The following parameters were measured and calculated: car- (Olympus, Japan).

Acta Pharmacologica Sinica www.chinaphar.com Feng GS et al 641

Western blotting assay Detection of SIRT-1 activity To extract the total protein, the mouse left ventricular tis - To detect the activity of SIRT-1, a SIRT-1 deacetylase fluoro- sue was homogenized in RIPA lysis buffer containing the metric assay kit (CycLex Ltd) was used. In this method, the protease inhibitor PMSF, and the total protein contents were reaction was initiated and the fluorescence intensity was mea- determined with a BCA protein assay kit (Pierce). To extract sured by mixing simultaneously fluorescence-labeled acety- the nucleoprotein, a nuclear extract kit (ACTIVE MOTIF) lated peptide (substrate), NAD+, and developer with SIRT-1 was used according to the manufacturer’s instructions, and samples. SIRT-1 samples were obtained in the immunopre- the nucleoprotein contents were determined with a Bradford cipitation assays using anti-SIRT-1 antibody. The fluorescence protein assay kit (Pierce). Then, equal amounts of protein intensity was measured at 490±10 nm (excitation) and 530±10 samples were separated by SDS-PAGE and then transferred nm (emitted light) by an automatic microplate reader (Tecan, to PVDF membranes (Millipore, MA, USA). After blocking Switzerland). with 5% (w/v) nonfat milk (Bio-Rad) dissolved in Tris-Tween 20 (TBS-T) for 1 h at room temperature, the membranes were Determination of NAD+ content incubated with anti-GAPDH, anti-lamin B1, anti-ANF, anti- The NAD+ levels were measured according to a method BNF, anti-β-MHC, anti-collagen I, anti-collagen III, anti-FN, described previously[22]. Briefly, the frozen crushed tissue was anti-PARP-1, anti-SIRT-1 or anti-PAR antibodies overnight at suspended in perchloric acid before centrifugation. The super-

4 °C. Then, horseradish peroxidase (HRP)-labeled secondary natant was neutralized with KOH and KH2PO4/K2HPO4, and antibodies were added for incubation for 1 h at room tem - then added to the NAD+ reaction mixture containing ethanol, perature. Immunoreactive bands were visualized with Super- 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, Signal West Pico Chemiluminescent Substrate (Pierce). The phenazine ethosulfate, EDTA, BSA, and bicine. The reaction intensity of target protein bands was quantified with Image J was initiated by alcohol dehydrogenase and stopped by iodo- software (National Institutes of Health, USA) and normalized acetate after incubation for 20 min at 37 °C. The absorbance of to GAPDH or Lamin B1 as loading control. the reaction mixture was read at a wavelength of 570 nm. The NAD+ content was calculated from the standard curve and Co-Immunoprecipitation normalized to the protein content of the sample. To investigate whether there was an interaction between SIRT-1 and PARP-1, a co-immunoprecipitation (co-IP) assay Statistical analysis was performed. Nuclear proteins (800 µg) were incubated Statistical analysis data are presented as the mean±SEM of with 3 µg anti-SIRT-1 or anti-PARP-1 antibodies overnight at least three independent experiments. Statistical analysis (rabbit normal IgG was used as a control), followed by a 4 h was performed using a two-tailed unpaired Student’s t-test incubation with protein G-agarose beads (Pierce, Rockford, IL, between two groups and by one-way ANOVA followed by USA) at 4 °C. The immunoprecipitated proteins were detected Bonferroni’s test for multiple comparisons. The analyses were by Western blot. performed using GraphPad Prism 5.0 (GraphPad Inc, La Jolla,

Scheme 1. Synthetic route of AG-690/11026014.

Acta Pharmacologica Sinica www.nature.com/aps 642 Feng GS et al

CA, USA). P<0.05 was considered statistically significant. alleviated the expression of the hypertrophic marker pro - teins (ANF, BNP and β-MHC) (Figure 3A–3C) and the fibro- Results sis marker proteins (collagen I, collagen III and FN) (Figure Synthesis of 6014 3D–3F), in contrast to Ang II. Taken together, these results To obtain a sufficient quantity of 6014 for the in vivo study, a suggested that 6014 could markedly prevent cardiac hypertro- novel synthetic route of 6014 was designed that is illustrated phy and fibrosis induced by Ang II. in Scheme 1. In brief, 6014 was synthesized through five steps, three of which could be completed simply and independently Effects of 6014 on the expression and enzymatic activity of on a column chromatograph to obtain pure product. This PARP-1 efficient synthetic route of 6014 was reported for the first time PARP-1 localizes mainly to the cell nucleus and to a lesser herein, and could be worthy for large scale synthesis. Finally, extent in the mitochondria[23–25]; therefore, the expression of an adequate amount of 6014 (2.0 g) was obtained and success- nuclear and total PARP-1 was detected by Western blot. As fully used for animal experiments. The purity and structure of shown in Figure 4A, 4B, both the nuclear and total expression 6014 were confirmed by 1H NMR, 13C NMR, HRMS and HPLC of PARP-1 was up-regulated upon stimulation of Ang II but (Figure S1–S4). was unaltered by 3-AB and 6014. Target proteins undergo the PARylation modification when PARP-1 is activated. Thus, the 6014 prevented cardiac dysfunction and changes of cardiac enzymatic activity of PARP-1 was evaluated by the expression structure in Ang II-infused mice of PARylated proteins using anti-PAR antibody. As shown To investigate the protective effect of 6014 on cardiac func - in Figure 4C, Ang II clearly stimulated the PARylation of pro- tion and structure, two-dimensional echocardiography was teins (3.20±0.30), whereas 6014 dose-dependently inhibited performed (Table S1). The echocardiography results indi - the increased expression of PARylated proteins (3.10±0.26, cated that the LVPWd, LVPWs, IVSd, IVSs were remarkably 1.53±0.06 and 1.20±0.10). increased, while the LVIDd and LVIDs were decreased in Ang Immunohistochemistry and immunofluorescence assays II-infused mice. The inhibitor 6014 dose-dependently reversed were subsequently performed to further confirm the inhibition these Ang II-induced changes, similarly as the well-known of 6014 on PARP-1 activity. Representative images of immu- PARP-1 inhibitor 3-AB (Figure 1A–1G). The LVEF, LVFS, SV nohistochemistry are shown in Figure 5A, 5B. Ang II clearly and CO have been most widely used to evaluate left ventricu- increased the expression of PARylated proteins, which were lar systolic function. Uniquely, LVFS was not significantly labeled with the anti-PAR antibody (brown). The up-regula- different among mice receiving different treatments (Table S1). tion of PARylated proteins was prominently inhibited by the Treatment with 6014 attenuated the Ang II-induced increases treatment of 6014 at the doses of 30 and 90 mg·kg-1·d-1, compa- of LVEF, SV and CO in a dose-dependent manner (Figure rable to 3-AB (30 mg·kg-1·d-1). Likewise, the fluorescence inten- 1H–1J). sity of PARylated proteins (red) was augmented in the hearts of Ang II-infused mice but was suppressed by 6014 (30 and 90 6014 attenuated cardiac hypertrophy and fibrosis induced by Ang II mg·kg-1·d-1) and 3-AB (30 mg·kg-1·d-1) (Figure 5C). Therefore, As shown in Figure 2A, Ang II treatment led to the enlarge- these observations were in line with the Western blot results ment of heart size, which could be prevented by 6014 at the (Figure 4C), confirming that 6014 efficiently inhibited the acti- doses of 30 and 90 mg·kg-1·d-1, as well as 3-AB at a dose of 30 vation of PARP-1 but did not affect its protein expression in mg·kg-1·d-1. Statistical results demonstrated that the ratios of Ang II-infused mice. HW/BW and LVW/BW were increased to 6.38±0.55 mg/g and 3.70±0.43 mg/g respectively by stimulation with Ang II, 6014 inhibited PARylation of SIRT-1 and restored the activity of while 6014 (and 10, 30 and 90 mg·kg-1·d-1) dose-dependently SIRT-1 in Ang II-infused mice inhibited the increase of HW/BW (5.97±0.55, 5.72±0.53 and Ang II treatment significantly decreased the NAD + content 5.46±0.61 mg/g) and LVW/BW (3.65±0.42, 3.26±0.23 and (80.3%±4.5%) and SIRT-1 activity (42.7%±5.0%) in left ven - 3.09±0.26 mg/g) (Figure 2D and 2E). HE staining images also tricles (Figure 6A and 6B). Treatment with 6014 did not retard showed that 6014 significantly prevented the enlargement of the decrease of NAD+ at 10 and 30 mg·kg-1·d-1 but did restore cardiomyocytes induced by Ang II (Figure 2B). the NAD+ content to 95.3%±5.0% of the sham group at the high Additionally, a significant increase of total collagen in the dose of 90 mg·kg-1·d-1 (Figure 6A), which may due to the strong hearts of the Ang II group was observed by the Masson’s inhibition of PARP-1 by 6014. SIRT-1 is an NAD+ dependent staining assays, which represented the appearance of cardiac enzyme that can compete with PARP-1 in NAD+ consumption. fibrosis (Figure 2C). Treatment with 6014 remarkably and Interestingly, 6014 could strongly restore SIRT-1 activity at dose-dependently prevented the accumulation of collagen the medium dose of 30 mg·kg-1·d-1 (66.3%±9.1%) compared to (10.3%±1.5%, 4.2%±0.76% and 3.0%±0.15%) compared to the the Ang II (42.7%±5.0%), even though the NAD+ content was Ang II group (14.4%±2.0%) (Figure 2C and 2F). The effect of not increased by 6014 at the same dose (Figure 6A and 6B). 6014 was comparable to that of 3-AB (6.4%±0.85%) (Figure 2C The results indicated that other factors underlying the SIRT-1 and 2F). activity regulation may exist in addition to the restoration of Western blot analysis showed that 6014 dose-dependently NAD+ content by 6014.

Acta Pharmacologica Sinica www.chinaphar.com Feng GS et al 643

Figure 1. 6014 reversed the cardiac dysfunction and changes of cardiac structure induced by Ang II. Representative echocardiographic graphs (A) were shown and IVSd (B), LVPWd (C), LVIDd (D), IVSs (E), LVPWs (F), LVIDs (G), LVEF (H), SV (I), and CO (J) were measured and calculated from parameters of echocardiography. Data were expressed as mean±SEM from three independent experiments. **P<0.01 vs Sham. ##P<0.01 vs Ang II.

Western blot analysis showed that 6014 dose-dependently tionally, Ang II enhanced this interaction significantly, and alleviated the expression of SIRT-1 in contrast to Ang II (Fig- 6014 attenuated the enhancement (Figure 6D). Moreover, the ure 6C), and the trend was consistent with that of SIRT-1 PARylation of SIRT-1 was detected using an anti-PAR anti- activity. SIRT-1 was precipitated by anti-PARP-1 (Figure 6D), body (Figure 6E). We observed two bands (approximately 170 suggesting an interaction between SIRT-1 and PARP-1. Addi- and 250 kDa), which were present at higher levels than the

Acta Pharmacologica Sinica www.nature.com/aps 644 Feng GS et al

Figure 2. 6014 attenuated the enlargement of the heart size, cardiomyocytes size and prevented the accumulation of collagen. (A) Representative diagram of mice hearts under different treatment. (B) Representative images diagram of H&E staining of cardiomyocytes. (C) Representative images of Masson’s staining of total collagen in tissue sections of heart. (D) Statistical results of HW/BW. (E) Statistical results of LVW/BW. (F) Statistical areas of total collagen production versus areas of whole tissue. Data were expressed as mean±SEM from three independent experiments. **P<0.01 vs Sham. #P<0.05, ##P<0.01 vs Ang II.

SIRT-1 band (120 kDa) (Figure 6E), suggesting the PARylation ment of the heart, i.e., cardiac hypertrophy. In the cardiac of SIRT-1. The results indicated 6014 significantly reduced the remodeling mouse model induced by Ang II, we observed PARylation of SIRT-1, in contrast to Ang II, which might con- that 6014 ameliorated cardiac hypertrophy by attenuating tribute to its activation of SIRT-1 activity. the enlargement of cardiomyocytes and the increase of HW/ BW and LVW/BW and by decreasing the expression of fetal Discussion genes, including ANF, BNP and β-MHC. Cardiac hypertro- In this study, we first designed a novel route of synthesis and phy, only when accompanied by fibrosis during remodeling, successfully synthesized a sufficient amount of 6014. During can lead to cardiac dysfunction [26]. It was also found that cardiac remodeling, the most apparent change is the enlarge- PARylation of PARP-1 is responsible for the pathogenesis

Acta Pharmacologica Sinica www.chinaphar.com Feng GS et al 645

Figure 3. Effects of 6014 on the expression of hypertrophy marker proteins and fibrosis marker proteins. Protein expression of ANF (A), BNP (B), β-MHC (C), FN (D), collagen I (E) and collagen III (F) were measured by Western blot. Data were expressed as mean±SEM from duplicates in three independent experiments. **P<0.01 vs Sham. #P<0.05, ##P<0.01 vs Ang II.

Acta Pharmacologica Sinica www.nature.com/aps 646 Feng GS et al

Figure 4. Effects of 6014 on the expression and enzymatic activity of PARP-1. The expression of PARP-1 in nucleus (A) and whole cell (B) as well as PARylated proteins in whole cell (C). Data were expressed as mean±SEM from duplicates in three independent experiments. *P<0.05, **P<0.01 vs Sham. ##P<0.01 vs Ang II.

of tissue fibrosis[27, 28]. Some researchers have reported that vivo. PARylation of c-Jun/c-Fos makes an important contribution Our previous work identified 6014 as a novel PARP-1 inhibi- to Ang II-induced cardiac fibrosis [29], and the inhibition of tor with obvious inhibition of the activity of PARP-1 at the PARP-1 prevented Ang II-induced aortic fibrosis in rats [30]. enzyme level and in a cardiomyocyte model. However, the Herein, we also found that 6014 prevented cardiac fibrosis underlying mechanisms of its cardiac protective effects in vivo induced by Ang II, as implied by the decrease of excess accu- remained unexplored. Herein, we further demonstrated that mulation of extracellular matrix (ECM) components, includ- the cardio-protective effects of 6014 depended on the inhibi- ing collagen I, collagen III and FN. Moreover, 6014 improved tion of PARylation, which was shown to activate SIRT-1. cardiac function, including LVEF, CO and SV and reversed SIRT-1, a prototypical member of the sirtuin family, is an the changes of cardiac structure in Ang II-infused mice. NAD+-dependent deacetylase enzyme[31]. Recently, SIRT-1 has Above all, these results confirmed for the first time that 6014 gathered significant interest due to its cardiovascular protec- exerts cardiac protective effects against cardiac remodeling. tive role[32, 33]. Several studies focusing on the responses of res- Importantly, 6014 showed better cardiac protective effects veratrol (RES) as an accepted SIRT-1 activator also indicated and lower cytotoxicity (Figure S5) than 3-AB, which proved that SIRT-1 may have a beneficial role in heart diseases and it to be very meaningful to investigate its cardio-protection in that the up-regulation of endogenous SIRT-1 is a protective

Acta Pharmacologica Sinica www.chinaphar.com Feng GS et al 647

Figure 5. Effects of 6014 on enzymatic activity of PARP-1. (A, B) Representative images of PARylated proteins expression by immunohistochemistry. PARylated proteins were stained with anti-PAR antibody (brown). (C) Representative images of PARylated proteins by immunofluorescence. PARylated proteins were stained with anti-PAR antibody (red) and nuclei were stained with DAPI (blue).

mechanism in cardiac diseases[34, 35]. Moreover, the cardio- of SIRT-1. In this study, overactivation of PARP-1 induced by protection of RES by activating SIRT-1 was also detected in Ang II in C57BL/6J mice did lead to significant consumption this paper (Figure S6, S7 and Table S2), which further con - of NAD+ and, subsequently, the down-regulation and inactiv- firmed the beneficial roles of SIRT-1 in heart diseases. Since ity of SIRT-1, as expected. More importantly, these pheno - both PARP-1 and SIRT-1 are present in the nuclear compart- types induced by Ang II were all reversed after the inhibition ment, the idea that they may compete for the common NAD+ of PARP-1 activity by 6014. We confirmed that the decrease substrate arose 10 years ago [36]. PARP-1 also has a higher of NAD+ consumption from the strong inhibition of PARP-1’s + affinity for NAD than does SIRT-1 (Km 20–60 μmol/L vs 100– activity by 6014 was one reason for reversing the activity of 300 μmol/L, respectively)[37, 38]. Therefore, over-activation of SIRT-1, which benefited the protection from cardiac remodel- PARP-1 can easily limit the NAD+ availability for the activity ing and dysfunction.

Acta Pharmacologica Sinica www.nature.com/aps 648 Feng GS et al

Figure 6. Effects of 6014 on the regulation and activity of SIRT-1. NAD+ content analysis (A) and SIRT-1 activity (B) were analyzed. Protein expression of SIRT-1 were measured by Western blot (C). The interaction between SIRT-1 and PARP-1 (D) and the PARylation of SIRT-1 (E) were measured by co-IP assay. Data were expressed as mean±SEM from three independent experiments. *P<0.05, **P<0.01 vs Sham. #P<0.05, ##P<0.01 vs Ang II.

Interestingly, we still observed the restorative effects of and PARP-1 in our later work. SIRT-1 in the case in which the activity of PARP-1 was clearly In summary, our study confirms that PARP-1 inhibition may inhibited but the NAD+ content was not reversed by 6014 at the suggest a therapeutic strategy for cardiac diseases. The novel dose of 30 mg·kg-1·d-1. Notably, the interconnection between PARP-1 inhibitor 6014 protected mice against Ang II-induced SIRT-1 and PARP-1 is more intricate[39], even though NAD+ cardiac remodeling, and ameliorated cardiac function. Treat- was reported as one accepted crosslink between them. Our ment with 6014 might be a potential therapeutic strategy for results also confirmed the direct interaction of SIRT-1 and heart diseases. PARP-1 by co-IP assay and, surprisingly, detected the expression of PARylation of SIRT-1, which was greatly up-regulated Acknowledgements by Ang II but down-regulated by 6014 at the dose of 30 mg· This work was supported by grants from the National Natural kg-1·d-1. Therefore, we deduced that the PARylation of SIRT-1 Science Foundation of China (No 81673433, No 81026548 and might contribute to the decrease of SIRT-1 activity, and the No 81273499 to Dr Pei-qing LIU), Team item of the Natural Sci- decreased PARylation of SIRT-1 by 6014 may be another new ence Foundation of Guangdong Province (No S2011030003190 reason for activating SIRT-1 under conditions of limited NAD+ to Dr Pei-qing LIU), the Major Project of Guangdong Province content; we aim to investigate the novel interaction of SIRT-1 (No 2015B020232009, No 2014B020210003, No 2013B090700010

Acta Pharmacologica Sinica www.chinaphar.com Feng GS et al 649

to Dr Pei-qing LIU), the Fund of the Guangdong Provincial cardiac fibroblasts. Cardiovasc Res 2009; 81: 98–107. Bureau of Traditional Chinese Medicine (No 20161049 to Dr 13 Pacher P, Liaudet L, Bai P, Virag L, Mabley JG, Hasko G , et al. Shao-rui CHEN), and the Medical Scientific Research Foun- Activation of poly(ADP-ribose) polymerase contributes to development dation of Guangdong Province (No A2015220 to Dr Shao-rui of doxorubicin-induced heart failure. J Pharmacol Exp Ther 2002; 300: 862–7. CHEN). 14 Pillai JB, Russell HM, Raman J, Jeevanandam V, Gupta MP. Increased expression of poly(ADP-ribose) polymerase-1 contributes to caspase- Author contribution independent myocyte cell death during heart failure. Am J Physiol Pei-qing LIU and Shao-rui CHEN defined the research theme Heart Circ Physiol 2005; 288: H486–96. and revised the manuscript critically; Guo-shuai FENG, Cui-ge 15 Bartha E, Solti I, Kereskai L, Lantos J, Plozer E, Magyar K, et al. PARP ZHU, Pan-xia WANG, Yi HUANG, Min LIU, Zhuo-ming LI, inhibition delays transition of hypertensive cardiopathy to heart failure in Ping HE, and Lan-lan LOU performed the research; Guo-shuai spontaneously hypertensive rats. Cardiovasc Res 2009; 83: 501–10. FENG, Cui-ge ZHU, Zhuo-ming LI, Shao-rui CHEN, and Pei- 16 Palfi A, Toth A, Hanto K, Deres P, Szabados E, Szereday ,Z et al. PARP inhibition prevents postinfarction myocardial remodeling and heart qing LIU analyzed the data; and Guo-shuai FENG wrote the failure via the protein kinase C/glycogen synthase kinase-3β pathway. paper. J Mol Cell Cardiol 2006; 41: 149–59. 17 Szabó G, Bährle S, Stumpf N, Sonnenberg K, Szabó É, Pacher P, et Supplementary information al. Poly (ADP-ribose) polymerase inhibition reduces reperfusion injury Supplementary files are available at the website of Acta after heart transplantation. Circ Res 2002; 90: 100–6. Pharmacologica Sinica. 18 Liu M, Li Z, Chen GW, Li ZM, Wang LP, Ye JT, et al. AG-690/11026014, a novel PARP-1 inhibitor, protects cardiomyocytes from Ang II-induced References hypertrophy. Mol Cell Endocrinol 2014; 392: 14–22. 1 Cohn JN, Ferrari R, Sharpe N. Cardiac remodeling-concepts and 19 Koh SH, Chang DI, Kim HT, Kim J, Kim MH, Kim KS, et al. Effect of clinical implications: a consensus paper from an international forum 3-aminobenzamide, PARP inhibitor, on matrix metalloproteinase-9 level on cardiac remodeling. Behalf of an International Forum on Cardiac in plasma and brain of ischemic stroke model. Toxicology 2005; 214: Remodeling. J Am Coll Cardiol 2000; 35: 569–82. 131–9. 2 Weber K, Brilla CG. Pathological hypertrophy and cardiac interstitium. 20 Nguewa PA, Fuertes MA, Cepeda V, Alonso C, Quevedo C, Soto M , Fibrosis and renin-angiotensin-aldosterone system. Circulation 1991; et al. Poly(ADP-ribose) polymerase-1 inhibitor 3-aminobenzamide 83: 1849–65. enhances apoptosis induction by platinum complexes in cisplatin- 3 Huentelman MJ, Grobe JL, Vazquez J, Stewart JM, Mecca AP, Katovich resistant tumor cells. Med Chem 2006; 2: 47–53. MJ, et al. Protection from angiotensin II-induced cardiac hypertrophy 21 Zhou SG, Zhou SF, Huang HQ, Chen JW, Huang M, Liu PQ. Pro and fibrosis by systemic lentiviral delivery of ACE2 in rats. Exp Physiol teomic analysis of hypertrophied myocardial protein patterns in reno 2005; 90: 783–90. vascularly hypertensive and spontaneously hypertensive rats. J Pro 4 Lim D-S, Lutucuta S, Bachireddy P, Youker K, Evans A, Entman M, et teome Res 2006; 5: 2901–8. al. Angiotensin II blockade reverses myocardial fibrosis in a transgenic 22 Yu SS, Cai Y, Ye JT, Pi RB, Chen SR, Liu PQ, et al. Sirtuin 6 protects mouse model of human hypertrophic cardiomyopathy. Circulation cardiomyocytes from hypertrophy in vitro via inhibition of NF-kappaB- 2001; 103: 789–91. dependent transcriptional activity. Br J Pharmacol 2013; 168: 117– 5 Paradis P, Dali-Youcef N, Paradis FW, Thibault G, Nemer M. Over 28. expression of angiotensin II type I receptor in cardiomyocytes induces 23 Scovassi AI. Mitochondrial poly(ADP-ribosylation): from old data to cardiac hypertrophy and remodeling. Proc Natl Acad Sci U S A 2000; new perspectives. FASEB J 2004; 18: 1487–8. 97: 931–6. 24 Du L, Zhang X, Han YY, Burke NA, Kochanek PM, Watkins SC , et + 6 Ichihara S, Senbonmatsu T, Price E, Ichiki T, Gaffney FA, Inagami al. Intra-mitochondrial poly(ADP-ribosylation) contributes to NAD T. Angiotensin II type 2 receptor is essential for left ventricular depletion and cell death induced by oxidative stress. J Biol Chem hypertrophy and cardiac fibrosis in chronic angiotensin II-induced 2003; 278: 18426–33. hypertension. Circulation 2001; 104: 346–51. 25 Luo X, Kraus WL. On PAR with PARP: cellular stress signaling through 7 Xu S, Bai P, Little PJ, Liu P. Poly (ADP-ribose) polymerase 1 (PARP1) in poly(ADP-ribose) and PARP-1. Genes Dev 2012; 26: 417–32. atherosclerosis: from molecular mechanisms to therapeutic implica 26 Grobe JL, Mecca AP, Lingis M, Shenoy V, Bolton TA, Machado JM , tions. Med Res Rev 2014; 34: 644–75. et al. Prevention of angiotensin II-induced cardiac remodeling by 8 Schreiber V, Dantzer F, Ame JC, De Murcia G. Poly (ADP-ribose): novel angiotensin-(1–7). Am J Physiol 2007; 292: H736–H42. functions for an old molecule. Nat Rev Mol Cell Biol 2006; 7: 517–28. 27 Mukhopadhyay P, Rajesh M, Cao Z, Horvath B, Park O, Wang H, et al. 9 Suzuki H, Quesada P, Farina B, Leone E. In vitro poly (ADP-ribosyl) Poly (ADP-ribose) polymerase-1 is a key mediator of liver inflammation ation of seminal ribonuclease. J Biol Chem 1986; 261: 6048–55. and fibrosis. Hepatology (Baltimore, Md) 2014; 59: 1998–2009. 10 Balakumar P, Singh M. Possible role of poly(ADP-ribose) polymerase 28 Hu B, Wu Z, Hergert P, Henke CA, Bitterman PB, Phan SH. Regulation in pathological and physiological cardiac hypertrophy. Methods Find of myofibroblast differentiation by poly(ADP-ribose) polymerase 1. Am Exp Clin Pharmacol 2006; 28: 683–9. J Pathol 2013; 182: 71–83. 11 Pillai JB, Gupta M, Rajamohan SB, Lang R, Raman J, Gupta MP. 29 Huang D, Wang Y, Yang C, Liao Y, Huang K. Angiotensin II promotes Poly (ADP-ribose) polymerase-1-deficient mice are protected from poly(ADP-ribosyl)ation of c-Jun/c-Fos in cardiac fibroblasts. J Mol Cell angiotensin II-induced cardiac hypertrophy. Am J Physiol Heart Circ Cardiol 2009; 46: 25–32. Physiol 2006; 291: H1545–H53. 30 Wang Y, Wang L, Zhang F, Zhang C, Deng S, Wang R, et al. Inhibition 12 Huang D, Yang C, Wang Y, Liao Y, Huang K. PARP-1 suppresses of PARP prevents angiotensin II-induced aortic fibrosis in rats. Int J adiponectin expression through poly (ADP-ribosyl) ation of PPARγ in Cardiol 2013; 167: 2285–93.

Acta Pharmacologica Sinica www.nature.com/aps 650 Feng GS et al

31 Yamamoto H, Schoonjans K, Auwerx J. Sirtuin functions in health and Cardiol 2016; 205: 99–110. disease. Mol Endocrinol 2007; 21: 1745–55. 36 Zhang J. Are poly(ADP-ribosyl)ation by PARP-1 and deacetylation by 32 Bindu S, Pillai VB, Gupta MP. Role of sirtuins in regulating patho Sir2 linked? Bioessays 2003; 25: 808–14. physiology of the heart. Trends Endocrinol Metab 2016; 27: 563–73. 37 Bai P, Cantó C, Oudart H, Brunyánszki A, Cen Y, Thomas C , et al. 33 Winnik S, Auwerx J, Sinclair DA, Matter CM. Protective effects of PARP-1 inhibition increases mitochondrial metabolism through SIRT1 sirtuins in cardiovascular diseases: from bench to bedside. Eur Heart activation. Cell Metab 2011; 13: 461–8. J 2015; 36: 3404–12. 38 Rajamohan SB, Pillai VB, Gupta M, Sundaresan NR, Birukov KG, 34 Thandapilly SJ, Wojciechowski P, Behbahani J, Louis XL, Yu L, Juric D, Samant S, et al. SIRT1 promotes cell survival under stress by deacetyl et al. Resveratrol prevents the development of pathological cardiac ation-dependent deactivation of poly (ADP-ribose) polymerase 1. Mol hypertrophy and contractile dysfunction in the SHR without lowering Cell Biol 2009; 29: 4116–29. blood pressure. Am J Hypertens 2010; 23: 192–6. 39 Xu S, Bai P, Little PJ, Liu P. Poly(ADP-ribose) polymerase 1 (PARP1) in 35 Cappetta D, Esposito G, Piegari E, Russo R, Ciuffreda LP, Rivellino A, atherosclerosis: from molecular mechanisms to therapeutic implications. et al. SIRT1 activation attenuates diastolic dysfunction by reducing Med Res Rev 2014; 34: 644–75. cardiac fibrosis in a model of anthracycline cardiomyopathy. Int J

Acta Pharmacologica Sinica